Twenty Product Images Cost $0.06 to $0.80 and a 30-Second Clip $1.50 to $9. The Hour You Spend on One Reference Image Is the Real Cost

Why this pays

A shop owner wants a listing dressed: twenty pictures of one product in the ratios the marketplace wants, and a thirty-second clip for the top of the page. A photographer is a day and a studio. From one approved reference image, the hosts' own price pages make the whole set for under ten dollars.

The host bill is not the cost. The hour you spend getting one reference image approved and writing down its seed is, and that hour is what makes the next product in the line take half as long. Everything below is how to spend it once.

The play, in three lines

1. Get one picture approved, a photograph or a first generation the client says yes to, and write down its seed and prompt.
2. In Replicate's FLUX 1.1 pro playground, make the twenty from that seed, changing only the prompt and the ratio; in fal's Wan 2.5 playground, animate three of them into 10-second clips and join them.
3. Save every file the same hour, put the seeds in your notes, and read the two licences on every image before it goes on a product page.

The path above walks every click. What follows is the reasoning behind each step and the numbers you quote from.

Where the money comes from

• A shop owner with a listing to dress. They pay for twenty files in the right ratios and a clip, all recognisably the same object, delivered this week.
• A print-on-demand margin, if the pictures sell a product you make. Then the eighty cents of images is the smallest line on the sheet, and Printify's base price, shipping and the marketplace's cut are the numbers that matter; the POD pages carry them.
• Not the host. fal sells prepaid credits and charges only for successful outputs; Replicate bills most models by run time, in arrears; OpenAI bills per second of video.

The pain you are selling the cure for

The listing needs one product on white, in a hand, on a table, in the room it belongs in, at 1:1 for the grid, 4:5 for the page and 16:9 for the banner, and every one must be the same mug. Twenty prompts give twenty mugs. One approved image with its seed written down gives one mug twenty times, and that consistency is the whole thing the client is paying for.

Step one: the reference image

• Start with one picture the client approves, a real photograph or a first generation iterated until yes.
• Replicate's FLUX 1.1 pro form has an image_prompt input, an image used together with the text prompt to guide the composition. fal's Wan 2.5 image-to-video takes an image as the first frame.
• One approved picture anchors everything else on this page.

Path A: the Playground tab

• Every Replicate model has a playground page with a web form. Open black-forest-labs/flux-1.1-pro: prompt; aspect_ratio from 1:1, 16:9, 3:2, 2:3, 4:5, 5:4, 9:16, 3:4, 4:3, or custom with width and height 256 to 1440 in multiples of 32; output_format webp by default, or jpg or png; output_quality 0 to 100, default 80; seed; safety_tolerance 1 to 6, default 2; prompt_upsampling, off by default.
• Set the format to png or jpg; some shop uploaders will not take webp. Run, wait, download. Web-interface predictions are kept indefinitely on your dashboard, with a Delete button.
• fal's model pages have a Playground tab generated from the input schema, with required fields marked, an upload box, a Run button, and each result kept on screen for comparison; the API tab shows the schemas, the page shows the price and the average latency. fal's Sandbox runs one prompt through several models side by side and shows the estimated cost before you submit.

Twenty images from one approved seed

• The schema line that makes a series possible: seed, set for reproducible generation. Write down the approved image's seed and prompt.
• Then the listing: the same product on a plain background, in a hand, on a table, in its room, at each ratio the marketplace wants. output_quality at 80 for the grid, 100 for the hero.
• Twenty runs of the form is an afternoon, which is why the API route exists.

Thirty seconds is three clips

• fal's Wan 2.5 image-to-video generates 5 or 10 seconds per run at 480p, 720p or 1080p, default 1080p. The first frame must be JPEG, JPG, PNG without alpha, BMP or WEBP, 360 to 2000 pixels a side, at most 10MB. Motion prompt up to 1500 characters; processing 1 to 3 minutes, longer for 10 seconds.
• Three 10-second runs from three approved stills, joined in any video editor with a cut or a crossfade. The page carries a Commercial use badge; the licence section below says why to read it anyway.

Path B: the API, for the twentieth listing

import fal_client
result = fal_client.subscribe("fal-ai/nano-banana-2", arguments={"prompt": "a futuristic cityscape at sunset", "aspect_ratio": "16:9"})
print(result["images"][0]["url"])

• Every fal Playground result includes generated code that reproduces the exact request, in Python, JavaScript and cURL. Swap the model id and arguments for the ones on your Wan or FLUX page.
• Replicate's equivalent is replicate.run("black-forest-labs/flux-schnell", input={"prompt": "A majestic lion"}), returning file objects you write to disk. Cold boots on a rarely used model can take minutes; only the time the prediction actually runs is charged.

Save the files the same hour

• Replicate: for predictions created through the API, all input parameters, output values, output files and logs are automatically removed after an hour by default. The script that makes two hundred images must write them to disk as it goes.
• Web-interface predictions are kept until you delete them.

A batch of twenty, and thirty seconds

The two charts above price this exact job at each published rate. Replicate: $0.04 an image on flux-1.1-pro, $0.025 on flux-dev, $3.00 per thousand on flux-schnell, so twenty images are $0.80, $0.50 or $0.06. Video, per second: fal's Wan 2.5 $0.05, Replicate's wan-2.1-i2v-480p $0.09, OpenAI's sora-2 $0.10 at 720p, sora-2-pro $0.30 at 720p, $0.50 at 1024p, $0.70 at 1080p, batch rates at half. Thirty seconds is $1.50, $2.70, $3.00 or $9.00. OpenAI prices its gpt-image models per million image tokens, gpt-image-1-mini at $8.00 per million output tokens, and points to a calculator instead of a per-image figure, so they are not on the image chart. The AI job cost calculator runs these for any batch size.

How each host bills

• fal: prepaid credits, drawn down as you use the platform, charged only for successful outputs, never for server errors or queue time.
• Replicate: most models by run time, setup and idle time free on public models, billed in arrears at the start of the month for the previous month; the first time usage crosses certain limits an account can be charged early as a fraud check.
• Together AI: pay as you go with non-refundable fees; credits funded by ACH or wire are not usable until the payment clears; access can be suspended on notice if any amount is 5 days past due. Its GPU prices, $3.99 an hour for an H100 or $1.99 preemptible, belong to the own-model page.

Who owns the mug's picture

• The host's licence: fal says the customer owns and retains all right, title and interest in the Customer Input, subject to a licence for fal to process it; Together says you exclusively own Your Content and Output and grant it only a limited licence to host them.
• The model's licence: Together notes the models may carry their own third-party terms; Replicate's FLUX 1.1 pro page says using it through Replicate means agreeing to the Black Forest Labs API agreement and terms of service.
• A picture going onto a product page needs both read. fal's Commercial use badge on the Wan page is the host's word, not the model owner's.

What expires, and what can vanish

• fal credits expire 365 days from purchase, promotional credits in 90 days, non-refundable and non-transferable; an account with no login or use for more than a year can be deactivated.
• Replicate may suspend or disable access to a Marketplace Model at any time for any reason, with no liability.
• A catalogue built on one model has one point of failure. Test a second model on the same reference image while the first still works, and keep its seeds too.

What to do today

1. Pick one product you already sell. Get one reference image approved.
2. Make five variations in the FLUX 1.1 pro playground with the seed written down, one 10-second Wan 2.5 clip from the best of them, and download everything.
3. Time it. The host bill will be under a dollar; the hour is the cost, and the seeds in your notes are what make the next product take half as long.